Air Pollution-Induced Autonomic Modulation

Abstract

Air pollutants pose a serious worldwide health hazard, causing respiratory and cardiovascular morbidity and mortality. Pollutants perturb the autonomic nervous system, whose function is critical to cardiopulmonary homeostasis. Recent studies suggest that pollutants can stimulate defensive sensory nerves within the cardiopulmonary system, thus providing a possible mechanism for pollutant-induced autonomic dysfunction. A better understanding of the mechanisms involved would likely improve the management and treatment of pollution-related disease.

Introduction

Air pollutants are a heterogeneous group of substances including particulate matter (PM), smoke, dust, ozone, sulfur dioxide, nitrogen dioxide, diisocyanates, aldehydes, etc. that are produced by human activity (e.g., combustion engine, plastics production) or naturally and (e.g., forest fires) that, when inhaled, are detrimental to human and animal health. Multiple epidemiological analyses (23, 24) show that inhalation of air pollution (in particular PM and O₃) causes significant clinical health impacts: for example, mortality risk rises 6% per 10 μg/m³ increase in PM_2.5 (≤2.5 μm) and, in 2005, ≤300,000 U.S. deaths were attributed to air pollution (56). Currently, worldwide data indicates that cities vary from 1 to 400 μg/m³ in PM_2.5, and climate change is predicted to cause these levels to rise (2, 54a).

Air pollutants dysregulate multiple biological processes, but the major causes of morbidity and mortality tend to be an increase in cardiovascular (CV) and respiratory events. Deriving mechanistic information from association studies is challenging, especially with multifactorial pollutant exposures that are likely chronic and fluctuating. Nevertheless, hourly increases in air pollutants have been associated with increased hospitalizations for serious CV events (19, 34, 80, 81, 100, 110, 123, 125–127, 151), even within 1–2 h (67, 97, 118, 129, 131, 133, 154). This rapid effect of pollutants on CV function suggests a significant neuronal component, in conjunction with other chronic effects (23, 24, 86). This review will provide 1) a brief overview of the evidence that pollutants cause serious health events via autonomic dysregulation, 2) a proposed mechanism, and 3) the current gaps in our knowledge.

Clinical Features of Pollution-Evoked CV Events

Inhalation of pollutants triggers ischemic heart disease and cardiac arrhythmia in susceptible individuals. The dysregulation of the autonomic nervous system (ANS) has been implicated in these events based on three lines of evidence. First, association studies suggest that pollutants cause tachycardia, hypertension, and reductions in the high-frequency domain of heart rate variability (HRV) together with cardiac arrhythmia, such as premature ventricular contractions (PVCs) in CV-diseased (CVD) patients with ischemic heart disease (12, 17, 38, 59, 65, 120) and hypertension (38, 120, 130), occupationally exposed workers (28, 35, 103), and the aged (30, 51, 64, 73, 82, 120, 144). HRV is a common tool in pollutant exposure studies for indirectly inferring autonomic balance. This analysis can derive a barrage of parameters that have been controversially interpreted as correlating with parasympathetic and sympathetic activity. In general, changes in the high-frequency domain (HiFq-HRV) in humans is strongly correlated with cardiovagal activity (i.e., parasympathetic stimulation) (71, 74, 137). However, it should be noted that muscarinic blockade modifies HiFq-HRV in a complex manner, HiFq-HRV is not abolished by vagotomy, and non-vagal factors can influence HiFq-HRV. Use of the low-frequency domain of HRV as a marker of sympathetic activity is even more controversial, and its mechanistic use is limited at best. Similarly, the ratio of low-frequency domain divided by HiFq-HRV is unlikely to be an appropriate marker of autonomic balance.

Second, pollution-evoked changes in CV parameters are rapid (28, 35, 51, 103), mimicking the acute effect of pollutants on CVD-associated hospitalizations. Third, association studies indicate a protective effect of β-adrenoceptor inhibitors on pollution-evoked CV events (59, 80, 100, 120). Thus it appears that inhalation of pollutants acutely shifts autonomic balance from sympathoinhibition to sympathoexcitation, which is a major risk factor for serious CV events (155). Interestingly, association studies (40, 79, 90, 130, 161, 166) and controlled exposure studies (8, 11, 51, 55, 60, 83, 121, 128, 134, 152) in healthy young individuals indicate that acute pollutants such as PM or O₃ have either no significant effect on heart rate, blood pressure, or HiFq-HRV, or cause bradycardia, hypotension, or increased HiFq-HRV [although an underpowered association study found differently (36)]. This suggests that pollutants cause distinct autonomic responses, depending on preexisting CVD: i.e., CVD individuals have sympathoexcitatory responses, and healthy individuals have sympathoinhibitory responses. Nevertheless, it is important to remember that autonomic nerve activity has yet to be directly measured in pollution exposures in any human population, thus caution is required in the interpretation of HRV, heart rate, and blood pressure data.

Animal Models of Pollutant Exposures

Most animal model studies investigate the biological effects of controlled and single-sourced pollutants, although some studies have placed animal cohorts in real-world environments such as roadways. The most common CVD animal models studied are the spontaneously hypertensive (SH) rat, and mouse and rat models of chronic heart failure caused by myocardial infarction following the permanent ligation of coronary arteries (MIHF). CV parameters are typically assessed using radiotelemetry in conscious animals. It should be emphasized, however, that the use of HRV to assess autonomic function in rodent models comes with many technical and theoretical caveats (74, 135). As with clinical HRV, changes in HiFq-HRV is partially representative of changes in parasympathetic activity, whereas the low-frequency domain of HRV is not a meaningful parameter of sympathetic activity. Pollutants including PM, O₃, and unsaturated aldehydes evoke tachycardia, hypertension, and decreased HiFq-HRV in SH rats and MIHF models (7, 26, 31, 32, 72, 122, 159, 162), whereas these pollutants cause bradycardia, hypotension or increased HRV in control/sham animals (7, 16, 37, 63, 69, 70, 92, 160, 162). As such, the animal model data are consistent with the clinical studies—pollutants cause sympathoexcitation in CVD animals but cause sympathoinhibition in healthy animals. That pollutants evoke autonomic reflexes is consistent with an increase in c-FOS immunoreactivity in subsets of neurons within the nucleus tractus solitarius (nTS) (see below) (62).

Mechanisms of Autonomic Cardiopulmonary Control: Afferent-Efferent Reflex

The importance of the ANS in the regulation of cardiovascular function has been known since the late 19th century. The two effector arms of the ANS, parasympathetic and sympathetic nerves, provide innervation to almost all peripheral tissues, including heart, lungs, and kidneys. The main neurotransmitters are acetylcholine and norepinephrine (and epinephrine) for the parasympathetic and sympathetic systems, respectively. Stimulation of the sympathetic system produces adrenoceptor-dependent physiological responses known as “fight or flight,” including increased cardiac output and blood pressure, dilation of the airways and pupils, and mobilization of energy stores, whereas stimulation of the parasympathetic system produces muscarinic receptor-dependent physiological responses known as “rest or digest,” including decreased cardiac output and blood pressure, constriction of the airways and pupils, stimulation of secretions and motility in the digestive tract, and deposition of energy stores. Excessive sympathetic function is a major risk factor for serious CV events in clinical populations (155).

Activity in preganglionic parasympathetic neurons in the medulla and in preganglionic sympathetic neurons in the spinal intermediolateral cell column is modulated by multiple neuronal and humeral factors. Sensory afferent activity (based on conditions within peripheral tissues), including from the spinal dorsal root ganglia (DRG) or cranial nerves such as the trigeminal, glossopharyngeal (via petrosal ganglia), and vagus, has profound effects on autonomic efferent activity. One well-characterized autonomic reflex that simply illustrates the major central autonomic pathways is the acute homeostatic control of blood pressure (46, 168): when blood pressure increases, baroreceptors in the carotid sinus and aortic arc activate glossopharyngeal and vagal afferents that terminate centrally in the nTS in the dorsal medulla. Activation of second-order nTS neurons leads to activation of 1) preganglionic parasympathetic in the nucleus ambiguus and 2) GABAergic neurons in the intermediate ventrolateral medulla (IVLM), which then inhibit presympathetic neurons in the rostral ventrolateral medulla (RVLM). Thus parasympathetic activity is increased and sympathetic outflow is decreased, leading to a drop in blood pressure due to resistance vessel vasodilation and decreased cardiac output (FIGURE 1), whereas a decrease in blood pressure causes a reduction in afferent activity, decreasing parasympathetic activity and releasing the inhibitory effect of IVLM neurons on RVLM neurons, thus resulting in a rebalancing of autonomic activity toward sympathoexcitation. Numerous other reflex arcs modulate autonomic CV networks (FIGURE 1), including the peripheral chemoreflex (168), Bezold-Jarisch reflex (149), pulmonary-cardiac reflex (42), and diving reflex (107, 119). Furthermore, many of these afferents directly modulate respiratory networks, which independently regulate autonomic activity (140, 168).

FIGURE 1.

Schematic showing the basic neural circuitry of CV reflexes evoked by afferent stimulation within the cardiopulmonary system

Excitatory afferent and central nerves are shown in green, inhibitory nerves are shown in red, parasympathetic nerves are shown in blue, and sympathetic nerves are shown in purple. ACh, acetylcholine; DRG, dorsal root ganglia; HR, heart rate; IML, intermediolateral cell column; IVLM, intermediate ventrolateral medulla; NA, nucleus ambiguus; NE, norepinephrine; nTS, nucleus tractus solitarius; RVLM, rostral ventrolateral medulla.

Airway Afferent Reflexes

The trachea, bronchi, and lungs are densely innervated by vagal afferent nerves, although there is also some DRG afferent innervation of the lungs. Despite our limited detailed knowledge of the vagal afferent subtypes innervating the human airways, numerous studies in animal models indicate that vagal airway afferents are heterogeneous with respect to size, myelination, gene expression, stimulus sensitivity, neurotransmitter expression, terminal location and structure, central connectivity, and function (33, 85, 89, 105, 116, 132, 156). The vagal ganglion is a combination of two distinct ganglia (nodose and jugular), and the embryological source of nodose afferents (placodes) and jugular afferents (neural crest) is a major determinant of gene expression and function (98, 105). The vast majority of vagal afferents have central projections to the nTS, with some afferent subtypes preferentially innervating specific subnuclei (33, 85) (FIGURE 1). A proportion of jugular afferents project to the paratrigeminal complex (85, 108), but little is known of their role in cardiovascular reflexes (53).

The distinct airway afferent subtypes have been recently reviewed (106, 157), so only the essential aspects will be included here. The majority of afferents innervating the airways are unmyelinated C-fibers. These are typically quiescent during eupnea but are activated by stimuli that are noxious or potentially noxious, and thus are termed “nociceptors,” according to Sherrington’s original definition (139). In general, nociceptor activation causes defensive reflexes that serve to protect the airways (e.g., cough, apnea, mucus secretion, bronchospasm) as well as help stabilize Pco₂ and Po₂ levels (e.g., bradycardia, hypotension) during the decrease in ventilation. The airways are also innervated by myelinated vagal A-fibers, including the well-characterized pulmonary stretch-sensitive mechanosensitive afferents, slowly adapting receptors (SARs), and rapidly adapting receptors (RARs), which provide feedback on lung volume, via the nTS, to respiratory centers in the medulla and pons (89). Nociceptive vagal C-fibers are activated by noxious stimuli due to their expression of a host of receptors and ion channels, which are activated by cellular damage, inflammatory mediators, irritants, toxins, oxidative stress, heat, acid, etc. SARs and RARs do not typically express such receptors and so are not directly activated by these stimuli, although mechanical changes in the lung caused by noxious stimuli (e.g., bronchospasm, edema) can activate these fibers indirectly.

A large majority of vagal C-fibers (from either nodose or jugular ganglion) innervating the airways are activated by capsaicin, the pungent ingredient in chili, due to their expression of the capsaicin-sensitive transient receptor potential (TRP) vanilloid 1 (V1) channel (27, 76, 87, 116). Many vagal C-fibers also express TRP ankyrin 1 (A1) (115), which is activated by a host of electrophilic irritants including allyl isothiocyanate (AITC; pungent ingredient in wasabi), cinnamaldehyde (pungent ingredient in cinnamon), H₂O₂, and other reactive oxygen species (ROS) (3, 10, 84, 142). Inhalation or intravenous injection of either capsaicin or TRPA1 agonists evokes defensive cardiopulmonary reflexes, including cough, apnea, hypotension, and bradycardia, in all mammals including humans (5, 20, 39, 41, 47, 54, 61, 77, 78, 94, 95, 117) (FIGURE 1). The hypotension and bradycardia evoked by these irritants are due to an increase in atropine-sensitive parasympathetic drive to the heart (FIGURE 1). TRPA1 and TRPV1 expression is largely limited in healthy postpartum mammals to nociceptive sensory subsets (27, 29, 84, 85, 113, 115), and neither are expressed in cardiomyocytes, efferent neurons, or carotid glomeruli (25, 88, 136, 143), although some reports suggest minor expression in other tissues (153).

For many years, respiratory neurophysiologists have used intravenous injection of chemical stimuli as a practical method for stimulating respiratory afferents. However, the right side of the heart is also innervated by vagal nociceptive C-fibers, whose activation also yields apnea, hypotension, and bradycardia. The CV portion of this response is termed the Bezold-Jarisch reflex (149), which is a cause of serious parasympathetic overactivation in some inferior wall myocardial infarctions. Many researchers considered the cardiac and pulmonary C-fibers that evoked these parasympathetic reflexes as being essentially identical. However, recent research by our laboratory suggests that irritant administrations via intravenous and inhalation trigger respiratory and CV reflexes that can be distinguished, particularly by their sensitivity to anesthetics (77, 78). It is currently not clear which specific nociceptive afferent subsets innervating the lower airways and heart are responsible for mediating the distinct parasympathetic reflexes evoked by irritants given by intravenous and inhalation. Injection of irritants into the systemic circulation (intra-arterially) invariably evokes hypertension and tachycardia, due to the activation of DRG nociceptive afferents (41, 43, 104). Thus vagal nociceptive afferents trigger sympathoinhibitory reflexes, whereas DRG nociceptive afferents trigger sympathoexcitatory reflexes.

Afferent Activation by Pollutants

We and others have consistently found that TRPA1 and TRPV1 mediate many of the biological responses to pollutants. PM (from diesel exhaust, coal fly ash, and heavy metal dust), PM components, acrolein, and other unsaturated aldehydes (e.g., formaldehyde and crotonaldehyde), isocyanates, and ozone have all been shown to selectively activate HEK293 cells expressing TRPA1 and nociceptive afferent neurons (typically vagal or trigeminal) (4, 15, 18, 20, 49, 50, 68, 91, 102, 109, 138, 145, 146). Furthermore, these responses are either dramatically reduced or abolished by selective inhibition or knockout of TRPA1 (FIGURE 2). Some PM have also been shown to activate HEK293 cells expressing TRPV1, and their activation of nociceptive neurons is reduced by selective TRPV1 inhibition (1, 48). In vivo responses to pollutants are largely consistent with the in vitro data: respiratory and CV reflexes evoked by pollutants are significantly reduced by selective inhibition or knockout of TRPA1 or TRPV1 (44, 63, 72, 92, 147). To a great extent, TRPA1 is activated by pollutants due to TRPA1’s sensitivity to electrophilic ligands, which is mediated through a number of highly reactive cysteines on the intracellular side of the channel (9, 75, 101) (FIGURE 2). The mechanism of activation of TRPV1 by some types of PM is less clearly understood but may be due to local acidification of the plasma membrane (1, 27). The majority of studies investigating the mechanisms of pollutant- and irritant-evoked reflexes were performed in healthy animals. Nonetheless, there is also evidence that TRPA1 mediates pollution-evoked autonomic dysfunction in CVD animals (72).

FIGURE 2.

Many pollutants activate nociceptive afferent nerves via the gating of the ion channel TRPA1

A: schematic showing the variety of pollutants/irritants that target reactive cysteines on the intracellular side of TRPA1, leading to its activation and cation influx. Examples of pollutants and irritants include (clockwise from top left) PM2.5, acrolein, toluene diisocyanate, formaldehyde, ozone (O₃), and hydrogen peroxide. B: calcium influx in HEK293 cells with or without expression of human TRPA1 (TRPA1-WT) or a mutant TRPA1 lacking critical cysteines (TRPA1-3CK) in response to AITC, diesel exhaust particles (DEP), ethanol extracts from DEP (DEP-EtOH), or coal fly ash PM (CFA1). *Significant response compared with empty HEK293 cells, # denotes significant decrease in TRPA1-3CK compared with TRPA1-WT (P < 0.05, ANOVA with Bonferroni correction). Adapted from Ref. 50, with permission from Chemical Research in Toxicology. C: calcium influx in dissociated trigeminal neurons from TRPA1^+/+ (i, n = 100 neurons) and TRPA1^−/− (ii, n = 100 neurons) mice in response to acrolein (100 μM), MO (AITC, 100 μM), and capsaicin (1 μM). Adapted from Ref. 15, with permission from Cell. D: action potential discharge in mouse bronchopulmonary C-fiber by O₃ (30 μM). Di, representative trace of action potential discharge; Dii, mean data showing action potential discharge + SD evoked by cinnamaldehyde (300 μM), O₃ (30 μM), and O₃ in the presence of the TRP channel blocker ruthenium red (RR, 30 μM) (n = 5). *Significant reduction by RR (P < 0.05, paired Student’s t-test). Adapted from Ref. 146, with permission from Journal of Physiology. E: toluene diisocyanate (TDI) evokes bradypnea and increased time of brake in wild-type and TRPA1^−/− mice. Ei, representative pulmonary airflow waveforms in response to vehicle and 1% TDI; Eii, mean respiratory rate in basal conditions and in response to vehicle (VHC) and 1% TDI in wild-type (black columns) and TRPA1^−/− (gray columns) mice (n = 7–10 for each genotype). *Significant decrease compared with vehicle (P < 0.05, ANOVA followed by Bonferroni posttest). #Significant difference between the mouse strains (P < 0.05, ANOVA followed by Bonferroni posttest). Adapted from Ref. 145, with permission from the American Thoracic Society.

Evidence for Remodeling of Irritant-Evoked Reflexes in CVD

The factors underlying the distinct autonomic responses to inhaled pollutants in CVD populations and animal models (sympathoexcitatory), and healthy humans and animals (sympathoinhibitory) is not well understood. We have recently begun to investigate such neuronal responses in a rat model of hypertension using AITC and capsaicin as model irritants that selectively activate TRPA1 and TRPV1, respectively. SH rats have polygenic arterial hypertension, which develops with age (starting at 7 wk) and is a risk factor for stroke and heart failure after >1 yr (52). We used SH rats at 15 wk of age, which have a mean arterial blood pressure (MABP) of 180 mmHg compared with 110 mmHg in age-matched normotensive WKY controls. Inhalation of either AITC or capsaicin by conscious SH rats evoked a complex brady-tachy arrhythmia accompanied by both AV block and premature ventricular contraction (PVCs) (78) (FIGURE 3). This was distinct from the irritant-evoked bradycardia with AV block in the WKY rats. When we analyzed each RR-interval (RRi), we found that AITC inhalation increased the percentage of tachycardic beats in the SH but not the WKY rats. The muscarinic inhibitor atropine abolished the AITC-evoked bradycardia in both WKY and SH rats, but atropine had no effect on AITC-evoked tachycardia or PVCs in the SH rat. Instead, the irritant-evoked tachyarrhythmia were abolished by the β₁-adrenergic inhibitor atenolol (FIGURE 3). These data suggest that AITC-evoked tachyarrhythmia in SH are due to de novo irritant-evoked sympathoexcitation. Importantly, the irritant-evoked tachyarrhythmia were abolished by vagotomy, indicating that vagal afferents were required. Furthermore, cardiac responses to electrical stimulation of vagal efferents is similar between SH and WKY (57, 58, 78), indicating that AITC-evoked tachyarrhythmia in SH are not due to aberrant cardiac responses to parasympathetic signaling. Again, we found evidence that irritants administered via inhalation and intravenously yielded distinct responses. Unlike inhalation, intravenous AITC evoked only bradycardia in SH rats (78, 163), suggesting that the reflex remodeling is restricted to only a subset of airway afferent pathways (FIGURE 3).

FIGURE 3.

CV reflexes evoked by inhalation of AITC are remodeled in hypertensive rats

A: representative ECG in normotensive WKY rats and hypertensive SH rats in response to vehicle and AITC. Note the bouts of tachycardia during AITC inhalation in the SH rat. The red asterisks denote PVCs. B: mean RR interval (time between each ventricular beat) + SD in response to vehicle and AITC in WKY (open columns, n = 9) and SH rats (red columns, n = 12). C: mean arterial blood pressure (MABP) + SD in response to vehicle and AITC in WKY (clear columns, n = 4) and SH rats (red columns, n = 7). D: effect of pretreatment with the muscarinic inhibitor atropine on AITC-evoked changes in RRi in WKY rats (Di) and SH rats (Dii). E: mean % of beats that are defined as tachycardic (+SD) in WKY and SH rats in response to either vehicle or AITC in control conditions (n = 9 WKY and 12 SH) or following pretreatment with muscarinic inhibitor atropine (blue, n = 6 WKY and 9 SH), β₁-adrenergic inhibitor atenolol (green, n = 5 SH), or angiotensin-converting enzyme inhibitor captopril (purple, n = 10 SH). F: mean number of PVCs + SD evoked by inhalation (Inh) or intravenous injection (IV) of either vehicle or AITC in WKY (black columns, n = 7) and SH (red columns, n = 9) rats. G: vagotomy abolishes the AITC-evoked decrease in RRi (+SD) in anesthetic-free, decerebrate, ventilated, and atropinized SH rats (n = 8). Statistics in B, C, E, and F: *significant effect of AITC; ^$significant difference between strains (P < 0.05, repeated-measures ANOVA with Bonferroni post-hoc testing); ^#significant effect of inhibitor pretreatment (P < 0.05, repeated-measures ANOVA with Bonferroni post-hoc testing). Statistics in G: *significant effect of AITC (P < 0.05, repeated-measures ANOVA with Bonferroni post-hoc testing); ^$significant effect of vagotomy (P < 0.05, repeated-measures ANOVA with Bonferroni post-hoc testing). All data are adapted from Ref. 78.

That irritant-evoked pulmonary-cardiac reflexes are remodeled in CVD is consistent with neuroplasticity in a number of other reflex pathways, including the baroreflex, the Bezold-Jarisch reflex, and the metaboreflex and the mechanoreflex from skeletal muscle (21, 66, 93, 96, 140, 150, 163). The cause of the remodeling of irritant-evoked pulmonary-cardiac reflexes in CVD is presently unknown. In the case of the SH rat, the steady-state hypertension is likely not a factor, since acute normalization of blood pressure using the anti-hypertensive captopril failed to abolish AITC-evoked tachyarrhythmia (78) (FIGURE 3). Thus it is likely the pulmonary-cardiac reflex is chronically remodeled in hypertension toward reflex sympathoexcitation. This could potentially be explained by the de novo recruitment of presympathetic neurons downstream of the activation of TRP-expressing afferents innervating the airways. Central respiratory networks are known to be remodeled in cardiovascular disease models; in particular, there is an amplified respiratory-sympathetic coupling in SH rats (22, 45, 114, 140). Furthermore, there is evidence that inflammatory signaling and neurotrophins (which are both associated with cardiovascular disease) can cause de novo expression of TRP channels in myelinated A-fibers innervating the airways (99, 165). More work is needed to determine whether these processes are critical to the observed remodeling of pollution-sensitive pulmonary-cardiac reflexes.

Chronic Stimulation of Stress Responses

Ambient pollution levels fluctuate, but for many individuals exposures are chronic. Chronic pollutant exposures cause inflammation, characterized by lung neutrophilia and edema, and systemic increases in oxidative stress, acute phase proteins, and inflammatory cytokines (23, 24, 63, 86, 111, 141). Chronic pollutant exposures have thus been implicated in vascular dysfunction, atherosclerosis, and glucose intolerance (13, 23, 24, 86). There is evidence that chronic activation of neuroendocrine responses is critical to some of the peripheral dysfunction caused by pollutants. Both ozone and particulate matter exposures increase plasma adrenocorticotropic hormone (ACTH) and corticosterone levels (148). In a set of detailed studies, ozone exposure for more than a few hours was shown to induce nTS neuronal activity, hyperglycemia, and glucose intolerance, as well as increases in circulating leptin and epinephrine (13, 62, 111, 141). The peripheral ozone-induced effects, along with lung inflammation, were reduced in rats with bilateral adrenal demedullation and were eliminated following bilateral adrenalectomy (112). It should be noted that, despite similarities in the inflammatory responses to pollutants, there are notable differences (14, 148). For example, ozone causes significant lung damage compared with particulate matter (158, 164), and this is reflected in the robust activation of antioxidant genes in the lungs by ozone only (148), whereas particulate matter inhalation causes much greater production of the xenobiotic-responsive cytochrome P450 family 1 member A1 enzyme in the heart than ozone (148).

The link between chronic pollution exposure and the activation of the neuroendocrine responses is not understood, although vagal afferents have been implicated (62, 86). Inflammatory pathways may also independently remodel reflexes (124, 167). More work is needed to understand the effect of chronic pollutant exposures on the remodeling of pulmonary-cardiac reflexes.

Summary

The impact of multiple pollutants on CV morbidity in association studies is well established, and evidence for acute autonomic dysfunction has been reported in both clinical populations and animal models. Autonomic responses to irritant and pollutant inhalation are modified by preexisting CVD, from sympathoinhibition toward sympathoexcitation, a major risk for CVD. Currently, the mechanism underlying the switch of pollutant-sensitive pulmonary-cardiac reflexes toward sympathoexcitation is unknown. There are also a number of other unresolved areas of investigation:

The effectiveness of β-blocker treatment in reducing pollution-related CV hospitalizations and mortality indicates the importance of pollutant-evoked sympathoexcitation. Nevertheless, how genetics contributes to susceptibility risk for pollution-evoked sympathoexcitation in CVD individuals is poorly understood.

Although mechanistic studies are possible in animal models, the evidence of afferent-mediated autonomic reflexes in clinical populations is only indirect (i.e., HRV studies). Improved autonomic assessment would be invaluable, as would electrophysiological recordings of vagal afferents and efferents. Furthermore, multiple studies have shown that irritant-evoked autonomic reflexes cause chaotic ECG responses, which diminishes the effectiveness of HRV assessment of autonomic function because chaotic beats are eliminated from these analysis.

Despite significant progress in the characterization of visceral afferents, the identity of the afferents activated by pollutant inhalation remains elusive. Although evidence favors subsets of vagal nociceptive C-fibers, the contribution of trigeminal, glossopharyngeal, and DRG afferents cannot be ruled out.

There are few clinical studies of the autonomic responses of each sole-sourced pollutant. Given the diversity of pollutants and their sources, one should be cautious to predict that all pollutants will act similarly, despite the similarities in their afferent activation profiles in vitro.